Dataset for the paper "When AI Navigates the Fog of War" (arXiv:2603.16642).
Website: war-forecast-arena.com
Overview
A temporally grounded benchmark for evaluating LLM reasoning during an ongoing geopolitical conflict. The dataset covers the early stages of the 2026 Middle East conflict, which unfolded after the training cutoff of current frontier models, substantially mitigating training-data leakage concerns.
Temporal Nodes… See the full description on the dataset page: https://huggingface.co/datasets/AIcell/war-test-dataset.
